I've been following this debacle in the German press and I think there's a structural problem that they will struggle to dig their way out of. And by "they" I don't mean just VW, but much of European industry.
Software is simply not valued in Europe, and not because there aren't amazing developers there -- there are many. But software isn't considered important
Let's start with cars. Back before cars were just computers with wheels I was briefly involved in a software project at $SERIOUS_GERMAN_CAR_COMPANY. Mechanically their cars were outstanding -- I still drive this company's cars today. This project was some cool "by wire" stuff, all modeled in Matlab, just as the ECU code was. But it was clear that the mechanical guys were the top of the pyramid; the safety guys were all mechanical engineers with some programming experience. All the user-visible electronics (radio, controls, etc) was subbed out to a low-price bidder because "who cares about that stuff anyway?". This wasn't VW, but I have some family exposure to VW specifically and that mentality still comes through deeply: electronics are added to the vehicle, not integral. The mental and organizational rewiring will be very hard. These are not companies who believe you should "eat your own lunch before someone else comes and eats it for you". The long institutional problems seen in yesterday's post about Nokia and the "Burning Platform" memo are pervasive throughout Europe (and most places, including a lot of USA): https://news.ycombinator.com/item?id=32698044
You see this in the salaries. Sure, take-home salaries in the EU are lower than US ones in across the board, but the delta in the high value professions is extreme and quite telling. My son's partner's parents in Europe are offended by what Amazon pays him in the US because "he just works in IT". Well, he's a developer in their highest revenue area, so Amazon think it's worth investing in. Tesla cars ship with all sorts of fit and finish bugs, and are above average in mechanical problems. But ("FSD" excepted) they spend more of their attention on what really matters: treating their vehicle as a modern electronic device. But the European car companies are still stuck in the mid 20th century and make the opposite branch cut. The rest of European industry is the same.
Andreeson wrote "Software is eating the world" 11 years ago. Apparently nobody on the continent of Europe has read it. Sure, software is considered important, but it's just another part of the BOM, not something strategic.
IMHO the only countries that really understand software at both a technical and business value add level are US, AUS, Canada, China, with India less so but in that group and Japan just barely getting in. Pretty damning.
FWIW I've worked in France and Germany (and non-European countries), including some car business, but most of my career has been in the Valley (starting 38 years ago). I am not from Europe or USA so in that sense I don't have a preference for either side. I prefer living in Europe but vastly prefer to work in the US.
Software is simply not valued in Europe, and not because there aren't amazing developers there -- there are many. But software isn't considered important
In many ways this is a self-fulfilling prophecy. An organization thinks software engineers aren't very important. And it doesn't pay the sort of salaries you need to get the top talent. So it doesn't have any top-tier software engineers, and their software team struggles to deliver valuable projects.
Later management changes their mind and decides to pay better salaries. They go to hire software engineers, and how do they even tell who the great software engineers are? Often they end up just paying more to the same not-so-great team, and watching as what they thought was an investment in software quality does not pay off.
I don't think this is specific to Europe, either - it can happen to any organization that hires a lot of inexpensive software engineers, in the US too. It's hard to get out of the "cheap software engineers" trap.
> In many ways this is a self-fulfilling prophecy. An organization thinks software engineers aren't very important. And it doesn't pay the sort of salaries you need to get the top talent. So it doesn't have any top-tier software engineers, and their software team struggles to deliver valuable projects.
And you can see this in their hiring. All full of interns. I remember when VW was advertising a “software school”. They were bragging about teaching dozens of “software engineers”.
The people who run these businesses think that writing software is like being a mechanic. That there’s some “end state” in software. These companies have no software culture. All that matters to them, and especially in Germany, is phds. They don’t care about software, they care that they employ a high number of phds.
Funny story. I remember attending an event at Telekom some years back. People from Fraunhofer ISST were on the panel discussing how to bring quality to a major European software initiative that is currently often on the news. Another point was to come up with an idea how to run an org that can maintain said sovereign platform in a transparent manner.
I asked them a question: why not to hand it over to the universities? Why not set up a number of working groups and use it as a platform to teach competent people on a real world, high profile system. You know: r&d, devops, monitoring, service, bug fixing, new features. Sounds like a great opportunity to teach future software generations, no?
You know what was the reaction of the ceo of Fraunhofer, the said man who was running that panel? He laughed and replied: “do you know how bad German universities are?” He said that in public in front of dozens of attendees. I was left speechless.
So if add 2 and 2, you understand what is happening. Everybody is happy with the status quo.
I mean, Fraunhofer is a major German research org and it should be in their business to promote such solutions. To fire something like this out, in front of the people who definitely attended those universities, that was shocking.
To this day I don’t know if I was simply naive or if my question was stupid in any way.
It sounds like you were in a meeting discussing how to set up a separate organization and your suggestion was to give the work to people who are doing a poor job already and then you are hoping that they'll suddenly stop failing if you give them access to private company details, work projects, even research. I think the mindset in these parts of Europe is that the companies are where you do real work and anyone who is actually good will find their way into a company anyway. Everyone who attends these universities knows that they are bad, it would not be a shock. They are there not because of the university but because of their own work ethic.
By and large, the 'best' research institutes in Germany are not the universities, but Max Planck and institutes like Fraunhofer. They may be loosely affiliated with the universities, but the University faculty and the research faculty are not one and the same. Students getting a doctorate may have that doctorate minted by the affiliated university, but all real work actually happens at the research institute.
Students can get doctorates directly from German universities and research is done there, but it's just not organized like American universities. Prime prestige centers are research institutes and they're spread around the country.
The alternative being? Another ERP system that does the same thing dressed up differently. Cobsidering how many comoanies world wide run SAP, it is fair to say that a true exonomic backbone software comes from Germany. If you take away SAP over night, we are all in a derp recession. If you you take away Facebook overbight, what would happen? Or Amazon, excluding AWS of course?
In Germany it is: a) 24 days vacation for 6 day working week, and b) 20 days for 5 day working week. That’s a minimum but there are also 11 days statutory national holidays in Germany (some Bundesländern have 12). The end result is a minimum of 31 paid out of work days if your work week is 5 days.
Europe is composed of a lot of countries, which all have different laws. In Romania, minimum vacation days are 20 days and very few companies give more than that, except maybe for seniority.
Is it really though? I ve been going around european companies and its all work till you drop 80h culture and no holiday, cause we need a guy on standbye. That stuff only ever exists on paper.
Which companies in which countries? I'm in Germany and I have never ever seen or (from friends and acquaintances) heard of vacations not taken due to pressure from the employer. It's about as common as an employer arbitrarily deciding to transfer less than the agreed amount of salary at the end of the month - maybe some very dubious employers have done it at some point, but it's not something to worry about.
What is fairly common is employers reminding employees to take their vacation days before they expire. Vacation days from year n expire in April or so of year n+1. That is intended to prevent postponing vacations to never (maybe until switching jobs).
Im in germany, and the pressure is there. Old employer at least gave vacations, but shifted them as needed. New employer basically denies vacations categorically, cause nobody else can take care of the system.
All this talk about labour laws etc. its like people hallucinating, cause in my world this just aint real.
Fraunhofer Society is a conglomerate of various research institutions. A lot of research in Germany is being done not at universities but at dedicated institutions like the Fraunhofer Society, the Max Planck Society, and many more.
The Fraunhofer CEO probably meant that the research part of German universities is bad, which is their main "competitor". This is still a ridiculous take and I attribute this to the general arrogance that was already mentioned by GP.
Yea, I am German myself and I really do not get this comment. I‘ve studied abroad and US universities / colleges are laughably easy in comparison. Grade inflation is a real thing and the result of inflated college fees.
Most US courses did not count as credit for my masters because they are too easy to get 100% in and they rarely cover even half of the average course load at TUM.
I have been involved in several comparisons of German to US universities. The good German ones are ahead of most/all US ones in terms of several academic measures.
Though most of those measures are about research output (# papers published, # high-impact papers, # awards won by faculty weighted by award prestige, # post-docs placed to tenure-track positions, # conferences chaired, # journal editors...)
I don't know. I come from a European university. When I was faced with US or UK students (I've been both an academic and a professional dev), I was scared by how little many of them actually knew. They were much better than us at communicating, much more self-assured, but were lacking other key skills.
:shrug:
That's one of the things that are really hard to compare, I guess.
I think the parent comment can be interpreted as "German universities are bad, but free" (the students don't pay tuition fees, or only a trivial amount)
aksss is being sarcastic. When universities are compared across nations, a frequent retort to the preeminence of the top US universities is "But universities in [insert European country here] are free!". aksss's point is that, as MrBuddyCasino indicated, you tend to get what you pay for.
US metrics based on research programs for how good a university is don't apply to Europe. Comparisons based on those are worthless and the fact it is still being done frequently can be attributed to either incompetence or malice.
Yes, it's so obviously a completely ridiculous metric that it has to be willful at this point. It doesn't seem like anyone tries to measure how well educated the undergraduates are in terms of skills acquired, etc, by how many papers a totally different population occupying a similar space pump out instead.
It shouldn't need to be said, but no amount of Nobel prizes in the world is going to make one whit of difference for introductory calculus instruction. When the prime research centers are decoupled from the universities and these silly little metrics, it's no wonder that European universities are "bad."
Or lack of a profit motive leading to not trying to game metrics as much. If an entity is ranked by some metric, and the better they are that metric, the more profitable they are, then they have an incentive to game it as much as possible.
Sure, but I would assume that the priorities for a private university are different than a public one. A private university might optimize for profitability and university ranking, while a public one might optimize for providing the most accessible and highest quality education while staying within budgetary guidelines.
It's like saying that American universities are the best on the "amount spent on American football facilities per student" ranking. It's true, but the other universities might not be competing on that particular metric and focusing on other ones.
The salaries at the big car companies are not bad actually (compared to other German companies... compared to the US they are pitiful, of course). The problem is the company culture. If you look at the recruitment material you will immediately notice this attitude of infantilization they have towards IT. They don't attract that kind of talent who's in it for the thrill of problem solving. They end up getting all the morons who like the brand and IT salary but have no passion for technology and it really shows in the product. Zalando is exactly the opposite: They sell extremely uncool shoes and are 1000 times as innovative as BMW with their sports cars. MOIA is also much better than their parent corporation Volkswagen.
I’ve seen big telco not having any internal dev - only bad management. When their partner do not deliver, or very slowly because they’re cheap, they make “escalation”…
So it’s even worse than not valuing the work. It’s destroying it.
(By the way, it’s not only telco, but also food distribution, small business…)
> But ("FSD" excepted) they spend more of their attention on what really matters: treating their vehicle as a modern electronic device.
Yet even Tesla does a rather bad job at this — they consider it to be a modern electronic device and ignore the wheels. OTA updates regularly change the UI in a way that is difficult and dangerous to learn when the wheels are spinning. Even when familiar with it, the very basics are hidden in menus and cannot be used without looking away from the road for longer than is safe.
Car companies: make your car software simple, reliable, and boring. You may well need to spend serious money doing this, but the goal should not be for it to be fancy or to feel extremely modern.
(With the latest update, turning off the seat heater is buried in a menu. I still haven’t figured out how to efficiently dismiss overlay apps that are covering the map. It used to be fairly obvious.)
It is well worth any Tesla driver's time to spend two minutes reading the list of voice commands. Once you have a grip on the vocabulary and structure you can just tell the car to do what you want for dozens of common tasks. (Hint: It's a "seat warmer", not a "butt heater".)
That's the good news, the bad news is that I still haven't figure out how to turn off the stupid cowbell when I accidentally bonk the right lever too many times. If I die in a car accident, 50-50 chance it will be because I was trying to stop the stupid cow bell. I wonder how often they run a query on post accident telemetry to see if the cowbell was on.
> It is well worth any Tesla driver's time to spend two minutes reading the list of voice commands.
It's not worth it. Not in any car. Voice commands are not a justification for shitty UX. Voice commands are not always applicable.
> Once you have a grip on the vocabulary
Why can't you see a problem with this? "Instead of using proper UX in the car, be a good trained monkey and learn these incantations that may or may not work"
“You need more cowbell. You gotta have more cowbell.”
That’s hilarious. I’ve never accidentally triggered it. But my son seemingly never gets tired of getting into the car early and turning on “fart on turn signal”.
Anything that gets him to finish getting ready and get happily settled in the car is a huge win in my book.
The idea that instead of turning a tactile dial I need to look away from the road to adjust temperature or volume if music, or I need to speak, is unacceptable to me. I consider ut terrible in terms of safety and UI.
My Tesla has two dials, one on each side of the steering wheel which allows for control of volume and temperature. There is never a need to look at the center console
In addition to the sibling coment, I also spend much more time distracted and listening to the response from the car to make sure it understood me that I would spend just turning a dial/pressing the button.
And I still have to glance at the screen to see that it did what I told it to do. Especially in the case of quite frequent failures to understand me.
They indeed are. One of the main reasons I paid good money for a Keychron Q6 keyboard was that it has a volume knob. So much better/faster than dealing with moving the mouse to volume icon and scrolling.
> IMHO the only countries that really understand software at both a technical and business value add level are US, AUS, Canada, China, with India less so but in that group and Japan just barely getting in.
What are Australia and Canada doing in there? Australia has… Atlassian? Besides that, I’ve only heard of developers in Aus/Canada/Japan being well paid if they work for US companies like Google or are parked there because of US visa issues.
Australians are well paid in general though, even if they don’t value software like FAANG does, because Australia is a petrostate pretending to be a service economy. Japan values software but doesn't pay anyone well because of the system where you earn salaryman virtue by suffering for the first 30 years of your career. (Hot takes.)
Australia has much more than just Atlassian. Canva is the 9th most valuable private company according to crunchbase [1].
But you need to look deeper than just "who has the biggest most successful companies. Note, the comment says "countries that understand software".
From what I've seen, and one of the reasons I moved here, Australia values quality production. They understand where software drives business and invest in that.
I'm former NICTA/CSIRO, and those organizations used to be drivers of technology and research.
Canada is a major technology hub. Microsoft and Facebook both have major development centres in Canada. Of course, being so close to the US, most of the brain drain goes directly out of the country, but that is part of the country and the people understanding the value of software. The education is good, and so people go to where the dollars are.
Canada and Australia are not going to be able to compete with the US from a dollar perspective. So many of the best Canadians, Australians, Chinese and Indian are going to end up in the US.
Canada has started to be competitive, mostly because of US companies coming in and bidding up wages, but when I left (~10 years ago) a top salary for a SWE was about $80k. I'm not sure it's because of some cultural appreciation of software as a value driver versus a cost center. The (self-proclaimed) branding as "Silicon Valley North" seems a bit excessive.
I also left roughly the same time as you, so I agree with what they were, but the range is truly wild now. You'll still find $80k positions but the ceiling is now easily 3-4x as much.
It's not the SV of the north yet -- local VCs are still a joke (expect one to two orders of magnitude worse terms) but it's getting better. The angel scene has actually been pretty incredible over the past few years.
Nice, then how comes everyone and their grandparents are running Jira and MS Office? The latter across FAANG as well (maybe not the G with an in house competitor).
This is because both companies sell to the enterprise, not to the end users.
This is how Microsoft drove their phone business off the cliff: they built a product that their customers (corporate IT) loved but the users detested. When the iphone showed up Microsoft famously knew it was dumb because it addressed nothing that their customers cared about. But the workforce demanded change.
I think you mean homegrown companies and Canada had Flickr, Slack, Autodesk, SideFX, Hootsuite, Shopify and Corel just to name a few off the top of my head.
Japan considers programming a blue collar task so pays accordingly (though there are exceptions, like Sony Computer Entertainment). But culturally the companies tend to understand the role of software better and have for a while. Remember how great the Keitei were in the late 90s/early 2000s compared to anywhere else in the world?
The kind of software in keitai (feature phones), car ECUs, and video games is different from the rest of the software industry. They’re good at making gadgets with embedded software and nothing else. In particular, nothing ever gets reused and nothing is ever made more abstract or flexible - that’s why most of the products can’t be exported and the iPhone killed all the Galapagos phones.
There are a lot of interesting lessons in video games though; it’s the only kind of software people love using enough to write emulators for, and game developers seem to exist in a lot more countries than other software businesses do.
There's a famous story in 2000s that Japanese feature phone engineer did hard work for low quality code. Phone manufacturers and telcos just wanted to make a great featured phone, but didn't focus on high quality software. https://lolipop-teru.ssl-lolipop.jp/gunsou/
Finally, they made very crappy Android phone and local industry shrunk. So Japanese consumers now love iPhone because Android phone is a trauma for some of them.
> What are Australia and Canada doing in there? Australia has… Atlassian?
There are plenty of wellknown tech/software companies originating in Australia. Xero, Freelancer, Canva, REA Group, MYOB, Airtasker, BigCommerce, Afterpay...
Many foreign software companies have development teams in Aus as well. Google, Zendesk, Square, Squarespace spring to mind.
I’m a US-UK dual citizen but my partner is Australian. I’m collecting the whole Anglosphere.
But when looking at living there, the wages/housing prices seemed like a poor trade off even compared to California. Although of course it’s a great place to live in many other ways.
I’ve heard of some of those but haven’t looked up their pay bands, not sure Afterpay has a moat though…
Having recently looked at wages/housing in the same anglosphere, Canada was way ahead of Australia. I know several folks who have left Australia for Canada in the last five years because of exactly this.
Maps originated in Australia, originally developed by an Australian company Where 2 Technologies employing two Danish brothers in Sydney.
> Google Maps first started as a C++ program designed by two Danish brothers, Lars and Jens Eilstrup Rasmussen, and Noel Gordon and Stephen Ma, at the Sydney-based company Where 2 Technologies. It was first designed to be separately downloaded by users, but the company later pitched the idea for a purely Web-based product to Google management, changing the method of distribution.[9] In October 2004, the company was acquired by Google Inc.[10] where it transformed into the web application Google Maps.
This is good insight. I does make me wonder if Tesla actually has a chance. Their software doesn't seem all that great. The UX in a Tesla got objectively worse on Christmas day, 2021 when they released a big change and gave us fart noises instead of a good UX. It now takes several more touches to do common tasks than it did December 24th, 2021.
For some inexplicable reason they have a web browser in the car. Every time I've tried it it crashes the Tesla's computer. Even viewing the Tesla manual reliably crashes the Tesla's computer.
If Apple really is working on a car, and if cars really are now just computers with wheels, then it's not at all clear that Tesla has the software skills to keep up with Apple (or whatever company actually has the skills).
I don’t think this phenomena is unique to Europe. I completed my degree in Mechanical Engineering, even though I was pretty sure near the end I was going to remain a programmer. The ability to bridge the two domains was appealing. So my plan was to exist in that place where software meets real world things. There have I worked for most of my career, the exception being a 6 year stint in tools/IDEs. The sentiment you describe has existed to some degree or another in all of my work. Software is always the “newcomer” to the game. Older electrical and mechanical hierarchies have usually come to pass. And so software ends up being managed as a “necessary evil”, by individuals who believe that their experience in other technical disciplines empowers them to manage and lead technical disciplines that are actually different.
> But ("FSD" excepted) they spend more of their attention on what really matters: treating their vehicle as a modern electronic device. But the European car companies are still stuck in the mid 20th century and make the opposite branch cut. The rest of European industry is the same.
You can probably say the same about GM or Ford, and I'd extend this to companies that sell hardware products in general. In the US too, the hardware department of many hardware companies runs the show, even though the hardware sold still runs a lot of custom software.
Tesla was shaped by a guy who got rich thanks to having founded a software company, and he brought those practices to hardware. It's the exception rather than the norm.
So the question should be, with Tesla and Rivian both being pretty young, where are the european startups? There is so much expertise in Germany for building cars, you would think that it is the best place to found a car startup. But where are they? DHL couldn't find a single German buyer for their car company.
I can’t speak to GM but from my experience with a Ford Mach E they’re starting to get it.
The interface, while not perfect, is not the kind of OEM junk I’ve seen from others. They’ve got working OTAs, although it’s going slow and they’re learning.
They may end up outsourcing a lot of that to Google’s Android based car platform (also called Android Auto?).
But they seemed to have learned from Tesla it’s an integral part of the car these days and not something you tack on at the end. It will take time but they are genuinely moving in the right direction at a reasonable (but not fast) pace.
> There is so much expertise in Germany for building cars, you would think that it is the best place to found a car startup. But where are they?
Precisely because they have the expertise they don't found startups. After founding an air company it's the second quickest way to become a millionaire if you are a billionaire. Tesla only became profitable in 2020 on the back of some government credits. 17 years after founding.
Remember Tesla's screens breaking because they used regular LCDs that couldn't withstand the vibrations and temperatures in a car? Yeah, that comes from expertise.
For a long time, there was a meme about "hardware" people not understanding software, and there was enough anecdotal evidence, to make the jokes about it funny. Example: "Never trust a hardware engineer with a keyboard, and never trust a programmer with a screwdriver."
I once mentioned this meme (referring to a specific anecdote) in a small group of CS people, and the big-name CS professor in the group immediately agreed with it, saying "They are our worst enemy!" (I think he was referring, seriously, to everything that has been learned, in his fields of CS and Software Engineering, beyond what needs to be exercised to get a program to compile and initially seem to work.)
My theory about that meme: all EEs (for example) will have done some programming (in some cases, like someone who can write numerous paragraphs of English, but isn't even aware of the gap between that and a great novel). And EEs might also reasonably have the impression that "coding" is easier than the hard real engineering classes that the EE had to pass. Of course, it's clear to everyone that CS people couldn't do the jobs of EE people, but it's less clear in the other direction.
To the extent that the meme is sometimes true, I think current market realities don't break it. A company driven by people with hard engineering/science backgrounds might realize it has to pay some market rate for software people. (And if the product obviously needed some kind of clown college skills, like face-painting, they'd pay for it, even if they thought those clowns had it easy.) But that doesn't mean they understand what's involved in the craft of effective software engineering, nor respect what goes into software staff's processes and judgment.
This isn't only about hardware and software people. I think it might be a more general human problem, across disciplines. It makes sense that software people will most notice the version in which they're on a receiving end, which I'd guess is usually on the other end of hardware people or (like most everyone else) business people.
(Additional theory: if an organization doesn't understand and respect some skillset, they might not be able to discern strong candidates when hiring for roles needing that skillset, and random results could be self-reinforcing.)
Of course a meme isn't truth, and it can be unfair to individuals. And some of the best software work I've seen was by people with EE/CE/math/no degrees, and I was also blessed early in career by working with great hardware engineers who did respect software.
The most complicated projects that humanity has built are software.
The tooling and the culture in the software world is decades ahead of fields like engineering or business management.
Take ideas like version control - a way to do change management. Open source. Test driven development. Continuous integration. These have existed in a nascent way in other fields, but software engineering has really taken the ideas to the next level.
I know some new hardware companies that have taken these approaches to their whole integrated hardware-software development methodology, with extreme success.
So it seems we need new car companies in Europe, and at least one founder probably needs to have used git.
> Take ideas like version control - a way to do change management. Test driven development. Continuous integration. These have existed in a nascent way in other fields, but software engineering has really taken the ideas to the next level.
TBF version control started (SCCS, RCS etc) by modeling how version control was done in mechanical engineering and architecture. Just look at the manual version control in old blueprints. Nowadays CAD systems like solidworks have a VCS for your models too. Absolute shit based on proprietary MS file formats, but it is there and it works.
OTOH there are reasons why you can't do TDD in the physical engineering, much less CI. When you're building a bridge you have to do FEA on your design, but the cost committment of construction doesn't allow much incremental change as you go -- almost all is in response to problems found once you've started building.
Code review is inherited from practices of other engineering disciplines.
At least in the dry conditions of South Australia there isn't a problem with rust ;-)
My old '99 model Porsche 911 996 is an exception to the rule. It's free motoring at this point; price rise since I purchased it 4 years ago has covered the cost of all and any expenses, including fuel and insurance.
Of course, the 996 was conceived in Porsche around the same time the Japanese came over and helped them out! They were in big trouble before they modernised.
Maybe this is a wake up call to the Germans to get in gear and sort themselves out before they run into (even bigger) trouble again!
>Now he reckoned that he has the most problems with European cars, most of which are considered premium or mid level brands in Australia.
Was this a surprise to you? Here in the US it's well known that European cars (including the expensive brands) are across the board worst in reliability, Japanese cars are the most reliable, and the US brands are in between.
I think you have just convinced me to buy German cars exclusively from now on; I very much do not want a car which is "just a computer with wheels" - the less software in the car, the better!
Oh you will still get a car with plenty of software, and the car will depend on it just as much as a Tesla. It's just that it's going to be badly made, buggy and not well integrated.
Those are going to suffer from electronics that flake out after a few years (flickering LED displays etc.) Long-term reliability is simply not a concern for German car manufacturers. You're supposed to buy a new car after a few years.
If you don’t want a computer on wheels, you don’t want a German car, you want a second hand Toyota Corolla.
We had an Audi Q5 first and a Tesla second. My wife and I fight over who gets to drive the Tesla.
With a modern German car, you still get some the fancy features, but they don’t work very well (lane control comes to mind), and the electronic UI is much worse than Tesla. CarPlay is the only way out.
And let’s not even talk about the difference in driving fun.
They're unreliable and some of them are quite offensive as well (BMW X6 types). If you don't want too much software, buy an Italian car. It's unreliable as well but it still limps with a shot sensor. And some of them are quite beautiful too. Otherwise I'd go for Japanese cars.
On the other hand, "German overengineering" (the mechanical side) is also a thing. Not as in built with huge margins and robustness, but intricate and overly complex.
Fun fact: There is no direct German translation for over-engineering.
EDIT: And now imagine what happens when the German drive to over engineer stuff, we do it with everything, not just hardware, meets software development for a complex system like a car... Then you get, well, VW.
I see this attitude in many domains, even among younger generations in startups where the majority of co-founders aren't developers, or after the original co-founders are sidelined by business people.
Software is just considered a necessary chore needed to do business, almost never a core part of the business with inherent value.
It's one of the reasons why Europe has almost no significant players in the IT space.
Out of curiosity, I just checked a rabdom Top 100 list. It seems to be market cap based, which in itself is a problem. What struck me is the absence of Dassault Systems for example while AutoDesk is in the top 50.
I do agree so that in Germany software as a stand alone thing has close to zero perceived value. If it is embedded in hardware, sure, it is important. Not as important as it could be, and it will always be seen as one part of many.
When it comes to stuff like MS Office, software is a tool, nothing else. As long as its working people a re happy using it.
The big devide, and I think it is a HN thing mostly, is how much importance software is given by people. Obviously, HN is giving it a lot.
Is this really a problem with European companies or more a problem with older companies which didn't need much software 30 years ago?
Are Ford and General Motors better at this than the European car manufacturers?
Maybe it is also caused by people staying very long at their employer in Europe. It is not uncommon to stay till retirement after you got into a big cooperation.
Well I followed the Berlin startup scene until I couldn’t bear it any more. Most of the “startups” were actually just small businesses that did business on the web. There’s absolutely nothing wrong with that, but a startup and a small business are different animals.
Ford and GM seem to have had a “Come to $DIETY” moment but yes, it remains to be seen if they will get their act together.
Ironically it was Mercedes who was the first to open an office in Palo Alto back in the mid 90s. It was considered absurd at the time, but all the other car companies now have offices in the Valley. But none of them are integral to their parent companies. Mercedes squandered their lead.
I find it interesting how the marketing depts. have woven together EV powertrains and “lots of electronic gizmos”. Nothing’s stopping an ICE car from having pop-out doorhandles, nor an EV with crank windows.
I don’t think marketing departments have anything to do with it. I just think Tesla built the first commercially successful EVs and they happened to let software control as much as possible. Now everyone building EVs is mimicking their preference for software, because its simply better. Anything that can be done by software, instead of hardware, should be. Hardware is difficult to change. Software can be updated and upgraded without a mechanic or a visit to a service station.
Great post! I would like to add, that Diess seemingly understood at least the strategic importance of this and invested heavily. The execution is (until now) just very bad. The German brands don’t “get” software the way e.g Tesla does. Funny enough, Musk said that Tesla is not even all that great in Software (exclude FSD). It’s just everybody else is so much worse. There is a great divide between them who feel software on a visceral level and those who don’t
I don't want software in my car being treated the same way andorid is on my phone. If everything is running, I want it as an embedded system that just runs. No OTA updates, air gapped to everything else, entertainment not connected in anyway to the running of the car.
Pay aside, I wonder if the German automotive industry is just a particularly bad example. I've worked in Germany, Switzerland, Canada and now for an American company and I found that software companies aren't not that much different.
I've also worked for an automotive supplier in Germany and can understand where parent is coming from.
Software companies have vastly lower overhead. Their staffing and cost structure is completely different from a manufacturer. SV tech bros getting $200K+RSU starting salaries can't be replicated by other businesses.
I had this experience in the US early in my software career. Supposed hardware companies didn’t see the value in software. As engineering leadership came up through hardware, they didn’t manage it well. Management pursued what they knew, and often the focus on hardware was at the expense of the user experience.
Now even in the US I think you see this with data. You see companies with managers brought up through software ranks that don’t understand how to measure and optimze systems and orgs with data (machine learning, but not only machine learning). They want to build complex software, often at the expense of the user experience, that just wants the thing that makes them think less…
The trouble is that everyone is clubbed together in "IT". Roboticists, distributed-systems experts, ML folks everyone is linguistically the same as someone who does IT support.
This being the case, it's not surprising that people don't understand the value of things: while they use Apple/Android etc. their immediate association with "IT" is that lowly dork who fixes their computer.
You see this in India too. From a industry full of low-level "IT-coolies" a few decades back, the aspirational places to work now are the "product" companies, where you don't just do grunt work.
The part about the mechanical guy being the top dogs seems more to do with the fact that a car must mechanically be safe above all else right? Like, I should be able to drive a car if the software fails. But the mechanical parts and components are the things that cannot fail or have a much lower tolerance to fail. It stems from a culture of not relying on software too much no?
There are few (no?) modern consumer vehicles that don't have safety-critical software somewhere and you couldn't legally drive any modern vehicles that don't have working software besides.
It's literally just a cultural issue that they think of themselves as car companies first, and good software culture isn't in their corporate DNA. 5-10 years ago literally no one understood that, but most executives are starting to get onboard with the need for changes and flail their way through them.
This is all completely true. EU salaries for programmers are a joke. In most of the EU, people still think programming consists in punching cards, following a pattern designed by somebody else (who is that person, nobody quite knows).
>IMHO the only countries that really understand software at both a technical and business value add level are US, AUS, Canada, China, with India less so but in that group and Japan just barely getting in. Pretty damning.
We're paid pretty low in Canada as well. The two devs that started at the same time I started my career had a salary of $11/hr ($7.62/hr USD), both of whom are quite talented. My salary when I stopped being doing testing directly was $52,000 at 9 years of experience. Toronto and Vancouver pay better but the cost of living there is quite high unless you already own the house you live in.
People should try to reconcile the facts that those outrageously high FAANG salaries occurs because of
1) Near monopoly in that sector
2) Excessive surveillance and tracking users
3) Increasing inequality between haves and have nots.
And non-FAANG type IT worker are also paid relatively high at the cost of other workers living perilous lives.
So unless Europeans really start cherishing these "values". They are not gonna get US level salaries. The way I see quite a lot non-US / European people just like high salaries but never want to think where it really come from.
All this talk about Europe does not respect/appreciate / values Software engineers sounds more and more like coded way to say We, software people, deserve much higher salaries at the cost of everyone else.
> those outrageously high FAANG salaries occurs because of ... 2) Excessive surveillance...
This is an absurd and lazy take.
There are plenty of high paying software jobs in the US that have nothing to do with surveillance (e.g. the AWS part of Amazon, Apple, SpaceX, all sorts of robotics companies, etc).
Europe's concerns in this area are a strength that they should be taking advantage of.
FWIW I think the surveillance itself is lazy. People just collect stuff because it's cheap to do, but the evidence that it is that useful in most cases is scant. If the FTC took a reasonable position that just having that info is a liability to be managed, most companies would stop and their businesses would do fine.
The point was, if I read it correctly, that comparing salaries is only part of the picture.
There is a yawning cultural divide between the USA and the rest of the world. Especially the non-English speaking world.
I am in an English speaking country that slips under the shadow of the cloud cast by the cultural hegemony of the USA (Aotearoa) and when I look around at my lifestyle, my friends, community, and the problems we have as a country and society - salary would not be enough to move me.
I am grossly underpaid by any standard (working on it) but the compensations of lifestyle and culture are well worth it
>I am in an English speaking country that slips under the shadow of the cloud cast by the cultural hegemony of the USA (Aotearoa) and when I look around at my lifestyle, my friends, community, and the problems we have as a country and society - salary would not be enough to move me.
There's nothing wrong with making such a decision. Many others make such a decision, whether in an English-speaking country or not. In the US itself, many make such a decision to stay in their home cities and states as opposed to moving to Silicon Valley.
But you also must recognize that you are in the minority. It is a fact that FAANG and comparable companies' Canadian offices are almost entirely staffed by non-Canadians that are parked there because they are either waiting for, or cannot receive, a US visa. Yes, there are some Canadians who want to stay close to home in Toronto or Vancouver, but they are very, very, very few.
Canadians have, thanks to treaties, a pretty simple path to working in the US. I think Australians also have a unique treaty in this regard. All other nationals have, to greater or lesser degree, various hurdles to clear before moving to the US. If tomorrow New Zealand signed a treaty with the US that made moving to the US for work as easy as it is for Canadians, how many of your coworkers and peers would stay?
> Reality is European workers are under-paid, not US workers over-paid.
My point was that European workers are paid in proportion to how they are valued by their employer. If they wrote transformational software their employer wouldn't understand its value and it would either just be shipped without a lot of thought, or ignored. Come to think of it this may be why so much great open source software comes out of Europe.
If European companies threatened by the digital transition realized their looming doom and actually decided to do something about it, the salaries would rise as a consequence.
The question is why and how. There is no point is arguing "low" European salaries unless someone shows European companies are more profitable and they save a nice chunk of money by paying less to S/w engineers.
Not very familiar with Australia. But thinking that high paying company like Atlassian which produces a very crappy software which I unfortunately deal daily.
To me it shows high pay does not guarantee good quality software which many(not you) seems to be arguing here.
> Not very familiar with Australia. But thinking that high paying company like Atlassian which produces a very crappy software which I unfortunately deal daily. To me it shows high pay does not guarantee good quality software which many(not you) seems to be arguing here.
That's a different problem. That's because you, the user, are not a customer of Atlassian. Companies and their micromanagers, however, love JIRA.
Microsoft is in the same boat. Virtually no users of their products are actually their customers.
That's how you get away with producing very crappy software, even if you hire the top talent.
> Microsoft is in the same boat. Virtually no users of their products are actually their customers.
That's how you get away with producing very crappy software, even if you hire the top talent.
Such a weird logic. You’re saying that Microsoft produces crappy software even though they hire too talent. And they get away with it because the people who buy it don’t use it (i.e. by extension don’t care if they’re spending their money wisely)?
> Apple has less than 10% of the PC market. What do you think the rest of them are running?
If you didn't pay for your copy of Windows (which you most likely didn't, far more likely it was provided to you in another way, such as OEM or your employer), you are not Microsoft's customer.
Likewise, if you got the free license from upgrading, you're also not their customer.
Pirated Windows? Not their customer.
Bought one of those $20 license keys? Still not their customer.
Actually bought a Windows key directly from them? Okay, you're their customer. You're in the minority.
I think it's a very Valley-centric opinion to think Microsoft doesn't have any consumer customers. Even putting aside all their other end user products consumers buy (XBox, the utterly dominant Office).
I certainly don't agree with your definition of a customer. People who buy a PC with a Windows OS on it know it's a Windows OS, and make that choice willingly. It's like pretending nobody is an Android customer because they buy their phones from Samsung not Google.
Perhaps GP means to highlight the distinction between the person deciding whether the money gets spent and the operation who uses the product. That can man a lot, eg if a boss just buys whatever he thinks it's good without asking the team what they want to use.
The point is that the majority of their revenue comes from corporate enterprise. This is one reason Ballmer didn’t understand the iPhone: his customers (corporate IT) wanted the windows phone but no actual users did.
Microsoft was taken by surprise by the “BYO movement”
> unless someone shows European companies are more profitable
They are. Most US "innovative competitive startups" are losing hundreds of millions, and often billions of dollars a year, for decades.
Europe (thankfully) doesn't want to do that. And "grossly underpaid software developers" enjoy a high quality of life on par with the rest of the population (universal healthcare, extended parental leave, vacations etc.)
> Most US "innovative competitive startups" are losing hundreds of millions, and often billions of dollars a year, for decades.
That isn’t actually true. A few certainly are, and they make great press fodder. But most don’t have that luxury. Actually it’s a burden: you suffer much more dilution.
Can you explain how paying software engineers higher salaries means other people need to live "perilous lives". You could just lower your EBIDTA a bit. Software engineers in the US aren't rich because of janitors being underpaid or something.
> And non-FAANG type IT worker are also paid relatively high at the cost of other workers living perilous lives.
Those workers live perilous lives because cities refuse to build enough housing, not because a few workers have it good.
> those outrageously high FAANG salaries occurs because of
4) Much more generous VC funding leading to more competition between firms for SWEs, as a result of a more risk seeking / risk tolerant culture with stronger work ethic
Also you're ignoring that even SWEs at companies like Walmart Labs make much more than SWEs in Europe.
> SWEs at companies like Walmart Labs make much more than SWEs in Europe.
Good. Find out how Walmart retail workers are doing. And Housing in US is lot cheaper for majority of people.
> ... risk tolerant culture with stronger work ethic
Yup, from Uber to Wework and dozen more food delivery service did awesome job with generous money in creating enduring businesses with strong work ethics.
> Good. Find out how Walmart retail workers are doing.
Plenty of other examples of well paid workers and US SWEs paid more than EU SWEs. UPS teamsters are unionized. AT&T technicians are unionized. Both pay their SWEs more than EU companies by a substantial margin.
> did awesome job with generous money in creating enduring businesses with strong work ethics.
You could name companies which are actually changing how (European) companies operate. Zoom, Slack, SpaceX, Figma, the list goes on and on. None of which were started in Europe.
32 hours a week and 60k EUR/year doesn't cut it in a competitive market, and any founder will tell you that.
> And Housing in US is lot cheaper for majority of people.
Not in any major city. Clearly you don't live in the US.
> Not in any major city. Clearly you don't live in the US.
Huh, lived in upscale DC suburb for many years, now moved to different city. Housing seems lot cheaper here when I checked with some friends based in some London suburbs. Another one with house in Amsterdam almost sounded same price for half size house than I bought. But yeah, no public transport for me and I am like 10 miles from downtown :(.
You're not serious? DC looks pretty expensive for an average worker, from a quick look at apartments.com. Housing is cheap in the South and in rural areas for sure, but any West Coast/northern East Coast city is very expensive. Suburbs used to be cheap but now are not so cheap after COVID+inflation.
Perhaps your expectations are calibrated based on European cities which also have housing shortages, like Berlin and London.
I think its very spread out at least in Europe in my experience. Hell even my own brother is envious on my salary because he got a engineering degree (vs my IT degree), so "by right" he should have a higher salary than me. Aside from the fact that he's employed by the local government to move paper from a desk to desk or rubber stamp documents.
My parents also think the same as being civil engineers themselves. Paraphrasing: doctors, lawyers and engineers should have the highest salaries whatever job they do, everyone else should just fight over leftovers.
Software is simply not valued in Europe, and not because there aren't amazing developers there -- there are many. But software isn't considered important
Let's start with cars. Back before cars were just computers with wheels I was briefly involved in a software project at $SERIOUS_GERMAN_CAR_COMPANY. Mechanically their cars were outstanding -- I still drive this company's cars today. This project was some cool "by wire" stuff, all modeled in Matlab, just as the ECU code was. But it was clear that the mechanical guys were the top of the pyramid; the safety guys were all mechanical engineers with some programming experience. All the user-visible electronics (radio, controls, etc) was subbed out to a low-price bidder because "who cares about that stuff anyway?". This wasn't VW, but I have some family exposure to VW specifically and that mentality still comes through deeply: electronics are added to the vehicle, not integral. The mental and organizational rewiring will be very hard. These are not companies who believe you should "eat your own lunch before someone else comes and eats it for you". The long institutional problems seen in yesterday's post about Nokia and the "Burning Platform" memo are pervasive throughout Europe (and most places, including a lot of USA): https://news.ycombinator.com/item?id=32698044
You see this in the salaries. Sure, take-home salaries in the EU are lower than US ones in across the board, but the delta in the high value professions is extreme and quite telling. My son's partner's parents in Europe are offended by what Amazon pays him in the US because "he just works in IT". Well, he's a developer in their highest revenue area, so Amazon think it's worth investing in. Tesla cars ship with all sorts of fit and finish bugs, and are above average in mechanical problems. But ("FSD" excepted) they spend more of their attention on what really matters: treating their vehicle as a modern electronic device. But the European car companies are still stuck in the mid 20th century and make the opposite branch cut. The rest of European industry is the same.
Andreeson wrote "Software is eating the world" 11 years ago. Apparently nobody on the continent of Europe has read it. Sure, software is considered important, but it's just another part of the BOM, not something strategic.
IMHO the only countries that really understand software at both a technical and business value add level are US, AUS, Canada, China, with India less so but in that group and Japan just barely getting in. Pretty damning.
FWIW I've worked in France and Germany (and non-European countries), including some car business, but most of my career has been in the Valley (starting 38 years ago). I am not from Europe or USA so in that sense I don't have a preference for either side. I prefer living in Europe but vastly prefer to work in the US.